In the case of shipper-carrier interactions there is ample econometric evidence that indicates that the interactions between shippers and carriers determine mode choice. This seems to be the result of the interactions by which shippers, after experimenting with various shipment sizes and receiving input (e.g., prices, level of service, damage rates) from the carriers, finally settle down on a given shipment size. The econometric evidence suggests that the shipment size choice, in essence, determines mode/vehicle choice. However, this econometric evidence, though compelling is not yet enough to prove the existence of a causal relationship between shippers decisions about shipment size and carriers mode choice decision. This paper provides insight into this issue by means of economic experiments. Data for testing the theory and for gaining insights about real decisions can be obtained by experimental economics where respondents will be playing shippers and carriers. In freight transport analysis, there are very few applications of experimental economics. The paper will describe the design, conduct and outcomes of an experiment carried out both at the Rensselaer Polytechnic Institute in the US and the Institute for Transport Studies of the University of Leeds. The objective of the experiment is to learn about interactions in the choice of mode and shipment size in freight transport between shippers as sender of goods and carriers as providers of transport services. In the initial set-up of the experiment with one shipper and several competing carriers, first the shipper decides on the number of deliveries in a week, which determines shipment size. To achieve this, the shipper minimises the sum of inventory cost (at the producing end) and the price paid for transport services of the carrier. Given the shippers choice on shipment size, each of the carriers then chooses a transport chain (characterised by the number of legs and mode/vehicle type for every leg), by minimising transport costs. The carrier then quotes a price for this shipment and the shipper selects a carrier. This is repeated for several rounds. Initially the shipper will choose relatively small shipment sizes (to keep inventory costs down) for which the transport costs per tonne will be high. The shipper has incomplete and uncertain information about the transport costs and has to learn about this from the offers of the carriers. In this situation the total logistics costs (including inventory and transport cost) will not be minimised, because the shipper only gets the transport cost information through the carrier. The experiment will show whether shipper-carrier interaction will lead to an overall optimum (and if so under which conditions) or whether the sub-optimality (too small shipments, and too high frequency) will be persistent. Several set-ups for the experiment are tested including experiments with a given market shipping rate, experiments where each carrier can choose from several modes and experiments where each carrier uses one particular mode. The outcomes of the experiments are analysed and reported in the paper. For the covering abstract see ITRD E137145.
Abstract